Kaggle cause-effect software
Python
Latest commit f4d0f0d Oct 15, 2015 Jose A. R. Fonollosa remove self in pipeline.predict()
Permalink
Failed to load latest commit information.
.gitignore
CEfinal_valid_pairs_head.csv
CEfinal_valid_predictions_head.csv
CEfinal_valid_publicinfo_head.csv
README.md
SETTINGS.json
data_io.py
estimator.py
features.py compatibility issues with new python modules Oct 15, 2015
hsic.py parallel Oct 1, 2013
predict.py Version 2.02 Oct 11, 2013
train.py
util.py

README.md

Copyright José A. R. Fonollosa jarfo@yahoo.com

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

   http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

code version: 2.02 code date: 09-OCT-2013 installation instructions: python 2.7 code. No installation required required python modules: numpy, pandas, sklearn, scipy Tested on a Linux machine (Fedora 17) with python 2.7.3 and the following versions of the python libraries numpy==1.6.2 pandas==0.10.0 scikit-learn==0.13.1 scipy==0.10.1

TRAINING (Aprox. 45 minutes)

  • Download train, SUP1 and SUP2 data from Kaggle (cause-effect competition)
  • Edit SETTINGS.json to indicate your data folders
  • Train the models: python train.py train train1 train2

FAST TEST (first 9 entries of the validation data)

  • python predict.py CEfinal_valid_pairs_head.csv CEfinal_valid_publicinfo_head.csv CEfinal_valid_predictions_head.csv Time to process: 10 seconds Results: (CEfinal_valid_predictions_head.csv rounded to 4 decimals places) SampleID,Target valid1, 0.70 valid2, 0.00 valid3, 0.00 valid4, 0.00 valid5, 0.00 valid6,-0.55 valid7,-0.14 valid8, 0.00 valid9,-0.01

See the data page of the Kaggle cause-effect competition for information about the data